Object Detection

R-CNN family, YOLO variants, DETR, anchor-free detectors, and open-vocabulary detection

1.

Selective Search WIP

Region proposal algorithm for object detection

2.

OverFeat

Sliding window + bbox regression + classification

3.

R-CNN Family

R-CNN, SPPNet, Fast R-CNN, Faster R-CNN evolution of region-based detectors

4.

YOLO

You Only Look Once — single-shot grid-based object detection

5.

SSD WIP

Single Shot MultiBox Detector

6.

RetinaNet

Focal loss for dense object detection addressing class imbalance

7.

CornerNet

Anchor-free detection using top-left and bottom-right corner heatmaps

8.

CenterNet

Center-based keypoint detection using corners as proposals

9.

ExtremeNet

Object detection via extreme point prediction

10.

DETR

End-to-end object detection with transformers using bipartite matching

11.

Swin Transformer WIP

Hierarchical vision transformer with shifted windows

12.

YOLOE WIP

Efficient YOLO variant

13.

YOLO-World

Open-vocabulary YOLO with vision-language modeling and RepVL-PAN

14.

DINO & Grounding DINO WIP

DETR with Improved deNoising anchOr boxes and open-set grounding

15.

OWLv2 WIP

Open-World Localization with vision-language models

16.

InternImage WIP

Large-scale vision foundation model with deformable convolutions